Antisense Therapies in Neurological Diseases

Jean-Baptiste Brunet de Courssou; Alexandra Durr; David Adams; Jean-Christophe Corvol; Louise-Laure Mariani


Brain. 2022;145(3):816-831. 

In This Article

Antisense Approaches

Molecular Structure of Antisense Therapies

Antisense therapies rely on antisense oligonucleotides (ASO), which are short, 12 to 25 nucleotides, single stranded DNA or RNA molecules, or double strand interfering RNA (iRNA) with several chemical modifications. They were first used for therapeutic purposes by inhibiting gene expression in 1978 by Zamecnik and Stephenson.[13] The chemical modifications are used to protect them against the action of nucleases that would otherwise degrade them too quickly, increase the strength of complementary hybridization and binding to proteins in the serum. They increase ASO and iRNA half-lives and allow the use of shorter oligonucleotides.

Concerning ASOs, their chemistry partly determines the mode of action: degradation of the target RNA, translation blockade or splicing modification. ASOs currently used to trigger RNAse H-mediated mRNA degradation all have a gapmer design, where a central sequence of deoxynucleotides is flanked by chemically modified oligonucleotides, increasing resistance to exonuclease-mediated degradation. Conversely, ASOs used for splicing modification or translation blockade do not need deoxynucleotides and therefore lack the gapmer design.

Most common chemical modifications in both ASO and iRNA (Table 1) are (i) formation of a bridged nucleic acid by linking the 2'-oxygen to the 4'-carbon of the ribose (locked nucleic acids); (ii) replacement of a non-bridging phosphodiester oxygen by sulphur, i.e. a phosphorothioate (PS) or phosphoramidate (PN) modification; (iii) complete replacement of the (deoxy)ribose backbone in phosphorodiamidate morpholino oligomer (PMO); and (iv) replacement of the 2'-hydroxyl by 2'-O-methyl (2'-OMe), 2'-O-methoxyethyl (2'-MOE), or 2'-fluoro.[14] Of note, when synthetized, both ASO and iRNA treatment are usually a mix of thousands of stereoisomers. However, controlling the stereochemistry is possible and could improve ASO efficiency and safety.[15–17]

Addressing Antisense Therapies to Target Cell Cytoplasm

Although the mutated gene is present in all the cells, it is usually expressed preferentially in specific cell subtypes where it exerts most of its deleterious effects, specific cells representing the ideal targets for antisense therapies. In most antisense therapies presented hereafter these target cells are neurons, with the noticeable exception of transthyretin-related hereditary amyloidosis and porphyria where the targeted cells are hepatocytes, and of myopathies such as Duchenne muscular dystrophy where the targeted cells are muscle cells. After intravenous delivery, ASOs accumulate in the liver and kidney, without renal excretion owing to their size. So far, they do not cross the blood–brain barrier, but intrathecal administration leads to ASO uptake by various cell types in the CNS, reaching the cytoplasm where they can exert their effects on the cytoplasmic mRNAs.[18] As the ASO distribution relies on CSF dynamics, uptake is heterogeneous in the brain parenchyma with a peak in superficial cortical regions.[19]

Double-strand RNA used for RNAi does not efficiently cross the cell membrane—requiring a continuous in situ infusion without vector—and is therefore delivered inside viral vectors—such as adeno-associated virus (AAV),[20] lipid shuttles or inside synthetic nanoparticles.[21–23] For viral therapies, antibodies against the vectors are classically exclusion criteria for the clinical studies and could constitute an obstacle for long-term treatment. For enhanced hepatocyte targeting, siRNA formulated into lipid nanoparticles have proved to be effective—as with patisiran in transthyretin-related hereditary amyloidosis.[24] Using N-acetyl galactosamine–siRNA (GalNAc-siRNA) conjugates is also effective, as with givosiran in porphyria.[25–28]

Molecular Mechanisms of Action of Antisense Therapies

Antisense Oligonucleotides Acting Through Splicing Modifications. Intronic regions are removed from the pre-mRNA sequence by the spliceosome enzymatic complex. Splicing may also remove exons in variable proportions, a process called alternative splicing, leading to different protein isoforms. This allows to increase the number of proteins encoded by one gene. Binding of ASO on key portions of the pre-mRNA sequence can hinder the normal splicing process, leading to intron conservation in the mature mRNA, or favouring an alternative splicing as described with nusinersen in spinal muscular atrophy (Figure 2A).[29]

Figure 2.

ASOs and iRNA mechanisms of action. (A) ASO fixation on precise regions of the pre-mRNA can alter the splicing, excluding some introns usually retained or favouring exon exclusion. (B) Interfering RNA is split into sense and antisense strand by the Dicer enzyme. The antisense strand is later loaded on the RNA-induced silencing complex (RISC) and binds to the target mRNA. Target mRNA is then degraded by the Argonaute 2 enzyme of RISC, and another mRNA can bind to the iRNA to be degraded. (C) ASO fixation on the mRNA lead, depending on the chemical structure of the ASO, either to mRNA degradation by RNAse H or to a hindrance of mRNA translation by the ribosomal complex. DMD = Duchenne muscular dystrophy; SMA = spinal muscular atrophy; TTRA = transthyretin-related hereditary amyloidosis.

RNA Interference. Different kinds of iRNA exist: small interfering RNA (siRNA), small hairpin RNA (shRNA) and microRNA (miRNA) (Table 1). iRNA targets mature mRNA. The choice for potential target nucleotide sequences is therefore narrower, as intron sequences are not considered.[30,31] The antisense strand of iRNAs will be loaded on the RNA-induced silencing complex (RISC; Figure 2B). The antisense strand binds to the target mRNA, which is degraded by the Argonaute 2 enzyme in RISC. RISC can degrade several mRNA as the iRNA antisense strand is preserved in this process and can pair repeatedly with complementary mRNA. Such an approach is used in transthyretin-related hereditary amyloidosis with Patisiran, an siRNA.

Antisense Oligonucleotides Acting Through mRNA Degradation by RNAse H. RNAse H degrades the RNA primer on the DNA, which is synthetized during replication. Pairing of a synthetic single-strand DNA ASO on mRNA will create a similar RNA–DNA pattern and activate RNAse H. The target mRNA is then degraded, while the chemically stabilized ASO is free and able to pair again to another mRNA (Figure 2C). These RNAse H ASOs are active in directing RNA cleavage in both the cytoplasm and nucleus.[32] ASOs currently used to trigger RNAse H-mediated mRNA degradation all have a gapmer design. For instance, Inotersen has a 5–10–5 gapmer design, meaning that the 10 central deoxynucleotides are flanked by five oligonucleotides, at the 5' and 3' end, and activates RNAse H in hepatocytes, used in transthyretin-related hereditary amyloidosis.[33]

Antisense Oligonucleotides Acting Through Translation Blockade. Some ASO, lacking the gapmer design, have the capacity to bind to mRNA without activating mRNA degradation. The ASO/mRNA pairing will preclude mRNA translation by the ribosomal enzymatic complex. This has the theoretical advantage of being an enzyme-independent process, avoiding enzymatic saturation and interference with the normal enzymatic processes of the cells (Figure 2C).[34] To the best of our knowledge, no drug is currently being investigated in an ongoing clinical trial with this approach in neurological diseases. Interestingly, targeting specific portions of mRNA with ASO such as upstream open reading frames could on the contrary be used to increase the translation efficiency of mRNA.[35]